Mixture-Of-Experts Collaboration for Intelligent OQA Management Ecosystem

Abstract

In the context of globalized market competition, product quality lies at the core of an enterprise's competitiveness. Among all aspects, Outgoing Quality Assurance (OQA) management serves as a crucial determinant in ensuring the quality of products. Traditional OQA management predominantly relies on manual inspections. This approach, however, suffers from low efficiency and is highly influenced by subjective factors, rendering it ill-equipped to address the intricate requirements of modern quality management.With the rapid advancements in technologies such as artificial intelligence and big data, the concept of intelligent systems has presented novel solutions for OQA management. This paper presents a multi-expert collaborative OQA intelligent management ecosystem. By integrating deep learning algorithms with expert knowledge, this management ecosystem enables real-time monitoring of product quality, intelligent early warning mechanisms, and precise classification of product defects. The management ecosystem is structured in a three-tier architecture. The data layer encompasses elements such as Lot running flow, Lot running history, and defect scanning data. The system layer incorporates modules for running early warning, defect management, and Case tracking. The application layer provides services including desktop systems, email notifications, and web-based interactive interfaces.This management ecosystem has demonstrated its efficacy in enhancing the timeliness of early warnings, improving the efficiency of Case handling, and increasing the accuracy of defect classification. As a result, it propels quality management towards greater intelligence and efficiency, offering a comprehensive OQA management solution tailored to the chip manufacturing industry.